bad for an amplifier when there is no speakers ?

Recently I had a friends new McIntosh Tube amp in my house (MC-2102) - at the time I was trying to break in a pair of interconnects which ran from my preamp to the amp. I disconnected the positive leg of the speaker cables from the amps binding posts and turned up the volume on my preamp to send a nice juicy signal through the interconnects - while I slept.

As I shut off the lights in the room. My attention was drawn to my system briefly as I swore I heard the music playing.

I crouched down next to the amp and listened.

Sure enough, I could hear the music coming from the amp itself without any speakers connected.

I then decided that it might not be a good thing to run an amplifier without the speakers connected since hearing IT make the music was quite disturbing.

The cause of this extends beyond my technical knowledge but I can speculate to presume the lack of an "output" for the amplifier caused the signal it was fed to feedback and hence become audible through the tubes.

Tube amps should NEVER be turned on without some type of load hooked up. Anyone that has done so has been lucky to say the least. I have seen a few output transformers that literally "melted down" from doing such.

The output section of a SS amp performs differently and would deal with this situation a little better. However, one should NOT have the amp on with ANY type of signal being passed with no load on it. Only BAD things can happen in such a case. This goes for both SS and tube gear.

If you are going to leave an amp on or try to burn in a cable without having to listen to the output, you need to at least connect a 8 - 16 ohm resistor across either the speaker leads or binding posts. This simulates a speaker load and protects the amp. Depending on the wattage of the resistor, you may be able to pass signal through it but i would not recommend advancing the volume control very high at all. This will act as what is known as a low power "dummy load".

The resistors that i'm going to mention are available from Parts Express ( www.partsexpress.com ). The part numbers for "generic" resistors that will handle up to 10 watts are 004-X. Substitute the actual impedance value for the "X". For instance, a 2 ohm would be 004-2, 004-8 would be an 8 ohm, 004-16 would be a 16 ohm, etc...

If you would like slighly higher powered resistors or "higher quality" models for use in a zobel network, try the 12 watt Mills. Part numbers work like the above but the values available are slightly different. Use 005-X with the resistance value being substituted for the "X". These are twice the price, handle slightly more power and are of tighter tolerance.

If you want to increase power handling, you can use multiple resistors in parralel or series, depending on the values that you choose. Just be careful to end up with a value that is suitable for your amplifier as a load. Going too low can stress the output devices and increase unnecessary heat.

Having said all of that, trying to burn interconnects in on a system is next to futile once you've heard the diferences of them being run on a Mobie or equivalent cable burner. Even with hundreds upon hundreds of hours on them, they would still not sound as relaxed or transparent as a cable that was run on a Mobie for just a few days. If you don't have access to a cable burner, try making friends with someone who has one and is an "overall nice guy" : ) Hope this helps.... Sean
>

My dad ran a tv and radio repair shop for many years in Brooklyn. As I remember one of his lessons ... it is ok to have the amp on, but potentially bad to have the gain (volume) passed 0. My dad had experiences with frying an amp with the power on and the gain up.

Eltbee , I spoke with an authorized NAD service tech a few years ago and told him I was having trouble with an NAD 214 that I was running in bridged mode. He said that in bridged mode excess stress is placed on the amp. During the course of that conversation he said that running only one channel of stereo amp actually reduced the load and would result in higher output while incurring less stress. I don't know if this statement is for amplifiers in general or if it only applies to the model in question, but it kind of makes sense to me.